CP and the Higgs A natural place to test for CP violating phases is - - PowerPoint PPT Presentation

CP VIOLATION IN h →τ+τ–

Roni Harnik, Adam Martin, Takemichi Okui, Reinard Primulando, FY

• Phys. Rev. D88 (2013) 076009 [arxiv: 1308.1094 [hep-ph]]
• U. of Massachusetts, Amherst, Amherst Center for Fundamental Interactions

The CP Nature of the Higgs Boson, May 2, 2015

Felix Yu Johannes Gutenberg University, Mainz

CP and the Higgs

• A natural place to test for CP violating phases is

with Higgs physics

– scalar-pseudoscalar admixture (e.g. scalar potential)

• naïvely tested via rate suppression

– couplings to gauge bosons (e.g. bosonic CPV)

• for example, tested via acoplanarity measurement in

h→ZZ*→4l

– couplings to fermions (e.g. fermionic CPV)

• our work: test via h → τ+ τ– → (ρ+ν) (ρ–ν) → (π+π0)ν (π–π0)ν
• [Full UV models to connect any given CP phase to a

baryogenesis mechanism is BTSOTW]

2

CP and the Higgs

• A natural place to test for CP violating phases is

with Higgs physics

– scalar-pseudoscalar admixture (e.g. scalar potential)

• naïvely tested via rate suppression

– couplings to gauge bosons (e.g. bosonic CPV)

• for example, tested via acoplanarity measurement in

h→ZZ*→4l

– couplings to fermions (e.g. fermionic CPV)

• our work: test via h → τ+ τ– → (ρ+ν) (ρ–ν) → (π+π0)ν (π–π0)ν
• [Full UV models to connect any given CP phase to a

baryogenesis mechanism is BTSOTW]

3

Outline

• Motivate new measurement in τ+τ– decay channel
• Sensitivity studies at colliders

– Lepton collider prospects – First proposal for an LHC measurement

• Summary

4

Testing “fermionic” CPV

• The BSM source of a CPV phase in SM Yukawa

couplings can be distinct from possible phases in the scalar potential or pseudoscalar couplings to gauge bosons

– Motivates CPV tests in fermionic couplings even if bosonic CPV coupling tests give null results – For example, new fermions which mix with SM fermions could introduce explicit phases in the Yukawa sector

5

Testing “fermionic” CPV with Higgs

• The tau decay channel for the Higgs is the most

promising system for direct measurement of fermionic CPV couplings

– Top coupling only probed via loops

• r ttH (tH) production

– Bottom quark polarizations generally washed out by QCD – Tau channel suffer from lost information via neutrinos (at hadron colliders), but still have an appreciable rate

MH = 126 GeV SM Br bb 56.1% WW* 23.1% gg 8.48% ττ 6.16% ZZ* 2.89% cc 2.83% γγ 0.228% Zγ 0.162% µµ 0.0214% 6

The h → τ+ τ– experimental status

• Both experiments have evidence and are actively

searching in all τ decay modes

7 CMS [1401.5041], ATLAS-CONF-2014-061

A Tau Yukawa CPV phase

• From an effective field theory perspective, can

readily generate a tau Yukawa phase via the addition of a dimension 6 operator

– α and β are generally complex – After inserting Higgs vevs, use the τR redefinition to get – Then, the Higgs coupling to taus is

8 Also see, e.g. Kearney, Pierce, Weiner [1207.7062]

A Tau Yukawa CPV phase

• The new phase can thus be captured by considering

the Lagrangian

– Δ = 0 is SM (CP-even) – Δ = π/2 is pure CP-odd (and CP conserving) – Δ = ±π/4 is maximally CP-violating – Δ is currently unconstrained (see next)

• We will assume the yτ magnitude is SM strength

9

EDM probe

• eEDM probes currently leave Δ unconstrained

10 Brod, Haisch, Zupan [1310.1385]

A CPV Observable

• We already lose

information from missing neutrinos

– Leptonic decays, though clean, lose even more information

• Need an intermediate

vector (not scalar) in the tau decay: focus on the ρ vector meson

– Br(τ+ → ρ+ ν) ≈ 26% – Br(ρ+ →π+ π0) ≈ 100%

11 PDG

Extracting the phase in Higgs decays

• Tau Yukawa CPV is imprinted on the tau

polarizations relative to each other

– Tau polarizations then get imprinted on the ν and ρ, ρ polarization is imparted to the πs

• Simplest observable (appropriate for LHC) is ρ+ρ–

acoplanarity angle

• New, better observable (appropriate for e+e–

collider) is Θ

12

Matrix element calculation

• Will trace how the CP phase Δ appears in the

squared matrix element by treating the Higgs decay as a sequence of on-shell 2-body decays

• Together, gives

13

Matrix element calculation assumptions

• Neglect π0 exchange (spatially separated; the τ’s are

boosted and back-to-back in the Higgs rest frame)

• All intermediate particles assumed on-shell
• Neglect π±–π0 mass difference
• Obtain

with

– Recall ρ± polarization is generally aligned with q±

14

Calculating the Theta Variable

• Introduce the variable

with coefficients

• We then write the squared matrix element as

where the most interesting piece is

15

Calculating the Theta Variable

• We can define an antisymmetric 2nd-rank tensor
• Or, even better, identify “electric” and “magnetic”

components

16

Calculating the Theta Variable

• We can calculate
• Specialize to Higgs rest frame (back-to-back taus)

– E+B+ and E-B- planes are parallel – Motivate a new acoplanarity between E+v+ and E-v- planes

17 Higgs rest frame

Ideal situation

18

Note MC Z background is flat

Ideal – compare to ρ+ρ- acoplanarity*

19 *Bower, Pierzchala, Was, Worek [hep-ph/0204292] Worek [hep-ph/0305082]

Θ amplitude is larger than φ* amplitude by 50%

Lepton collider possibilities

• We obviously cannot directly measure neutrino

momenta

• At a lepton collider, have enough constraints to

solve algebraically for neutrino momenta

– Have two neutrino momenta solution sets

• Both solutions give correct Higgs mass
• Weight each solution by half an event
• Necessarily require visible Z decay
• Higgs events tagged via recoil mass

20 ILC TDR Volume 2

Lepton collider – reconstructed

21

Reconstructed amplitude degraded by 30%

Lepton collider – reconstructed

22

Lepton collider possibilities

• For √s = 250 GeV ILC, polarized beams, Zh

production is about 0.30 pb

• With unpolarized beams (FCC-ee or CEPC), cross

section is about 30% less

• ILC signal yield (using SM Br(h →ττ) and restricting

to visible Z decays) is 990 events with 1 ab-1 luminosity

23

Lepton collider possibilities

• For √s = 250 GeV ILC, polarized beams, Zh

production is about 0.30 pb

– ILC signal yield (using SM Br(h →ττ) and restricting to visible Z decays) is 990 events with 1 ab-1 – Construct binned likelihood using a sinuisoidal fit to signal, determine sensitivity by variation of test Δ

24

With 1 ab-1 of ILC √s=250 GeV, expect 1σ discrimination of 4.4° (compared* to 6° using φ* [albeit included backgrounds and detector effects])

*Desch, Imhof, Was, Worek [hep-ph/0307331]

Luminosity scaling (without systematics)

25 (1.15 degrees) (5.7 degrees) CEPC or FCC-ee lum. is 30% smaller

Luminosity scaling (without systematics)

26 (1.2 degrees) (5.7 degrees)

With 10 ab-1 of FCCee or CEPC √s=250 GeV, expect 1σ discrimination of 1.7°

Lepton Collider Prospects

• Systematics will affect high luminosity estimates
• Expect some minor sensitivity losses from detector

resolution

– Z recoil mass with ee and μμ resolution is highly superior to other channels

27

ILC (1 ab-1) FCCee/CEPC (ab-1)

LHC prospects

• Consider h+j events (“boosted” τhadτhad sample)
• At the LHC, need to approximate neutrino momenta

– Have (8-2-2-2=) 2 unknown four-momentum components – Will use collinear approximation for neutrino momenta

• In this approximation, Θ is identical to ρρ acoplanarity angle
• Other approximations considered tended to wash out or

distort the sinuisoidal shape of the Θ distribution

– First proposal to measure Δ at the LHC with prompt tau decays and kinematics

28

Ideal vs. Collinear approximation

29

Collinear amplitude is about 25% of the truth Θ amplitude

LHC14 simulation details

• Use MadGraph5 for h+j and Z+j events at LHC14

– Mimic cuts for 1-jet, hadronic taus Higgs search category – Impose preselection of pT(j) > 140 GeV, |η(j)| < 2.5 – Normalize to MCFM NLO ς(h+j)=2.0 pb, ς(Z+j)=420 pb – No pileup or detector simulation, aside from tau-tagging efficiencies

• Pileup degrades primary vertex determination for charged pion

tracks and adds ECAL deposits that reduce neutral pion resolution

• Tracking and detector resolution will clearly smear the Θ

distribution

30

Yields for 3 ab-1 LHC

• Signal region:

MET > 40 GeV, pT(ρ) > 45 GeV, |η(ρ)| < 2.1, mcoll > 120 GeV

– Inject an additional 10% contribution to (flat) Zj background to account for QCD multijets

31

Nevents for 3 ab-1 with τ-tagging 50% efficiency

Yields for 3 ab-1 LHC

• Consider τ tagging efficiency benchmarks of 50%

and 70%, use likelihood analysis testing different Δ

– Discriminating pure scalar vs. pure pseudoscalar at 3σ requires 550 (300) fb-1 with 50% (70%) τ tagging efficiency – For 5σ, require 1500 (700) fb-1 with 50% (70%) τ tagging efficiency

• Again, detector effects and pileup are neglected

32

Luminosity scaling (without systematics)

33 (8 degrees) (4.6 degrees) (17 degrees)

Improving the measurement of the tau Yukawa CP phase

• Consider including MET information for LHC

analyses

– e.g. MELA-type likelihood incorporating signal hypotheses with different Δ

• Consider other tau decay modes or add decay

vertex information

• Improve tau tagging efficiency
• Dedicated di-tau hadronic trigger
• Consider VBF production, Zh production

– For VBF, 3 ab-1, expect 52k π+π0ν π–π0ν total events (no cuts)

• S/B is about 0.4 from ATLAS 8 TeV BDT analysis

34

Recent Delphes analysis*

– Collinear approx. at LHC is likely a hard limit – Angular resolution is negligibly (4%) degrades Θ distribution – Energy resolution affects contamination from irreducible Z background

35 *Askew, Jaiswal, Okui, Prosper, Sato [1501.03156]

Summary

• New CP phases are motivated from general baryogenesis

arguments

• Have a new suite of measurements to perform in Higgs

physics – Fermionic CP phases play a special role – Look forward to implementing this analysis in future Higgs studies – Can also consider prospects at FCC-hh and SPPC

36

Motivating CPV tests

• Sakharov’s three conditions for baryogenesis

motivate searches for new sources of CP violation

– Need B violation – Need C and CP violation – Need interactions to happen out of thermal equilibrium

• Our picture of baryogenesis is currently incomplete

– SM EW baryogenesis is insufficient – Should probe for new sources of CPV

38

CP and the Higgs

• A natural place to test for CP violating phases is

with Higgs physics

– scalar-pseudoscalar admixture (e.g. scalar potential)

• naïvely tested via rate suppression

– couplings to gauge bosons (e.g. bosonic CPV)

• for example, tested via acoplanarity measurement in

h→ZZ*→4l

– couplings to fermions (e.g. fermionic CPV)

• our work: test via h → τ+ τ– → (ρ+ν) (ρ–ν) → (π+π0)ν (π–π0)ν
• [Full UV models to connect any given CP phase to a

baryogenesis mechanism is BTSOTW]

39

UV completion

40

Yields for 3 ab-1 LHC

41 Green = SM signal Red = pseudoscalar signal Purple = Cosine fit to SM Blue = Cosine fit to pseudoscalar

For lower luminosity, the amplitude is the same but the significance of a non- zero phase shift is less

Tau measurement details

• Method relies on reconstructing neutral and

charged pions with good resolution and efficiency

CMS JINST 7, P01001 (2012) [arXiv:1109.6034 [physics.ins-det]] 42

Measuring Higgs to ττ

• Use SVFit to reconstruct mττ (creates likelihood

function based on observed kinematics)

– Anticipating the CP phase measurement, focus on the fully hadronic analysis

43 CMS PAS-HIG-13-004

Measuring Higgs to ττ

• Use SVFit to reconstruct mττ (creates likelihood

function based on observed kinematics)

– Anticipating the CP phase measurement, focus on the fully hadronic analysis

44 CMS PAS-HIG-13-004 Combined: μ = 1.1 ± 0.4

ATLAS Update

• Use BDT output to categorize events

45 ATLAS-CONF-2013-108

ATLAS Update

• Use BDT output to categorize events

46 ATLAS-CONF-2013-108

ATLAS Update

• Focus on fully hadronic channel

– Main backgrounds are still irreducible Z →ττ and QCD multijets

47 ATLAS-CONF-2013-108

Tau measurement details

CMS JINST 7, P01001 (2012) [arXiv:1109.6034 [physics.ins-det]] 48

Tau measurement details

CMS JINST 7, P01001 (2012) [arXiv:1109.6034 [physics.ins-det]] 49